This proposal has four major divisions: 1. The continuation of computational support for the molecular biology research community. 2. The dissemination of newly developed computational technologies and application examples. 3.The continuation of a multi-level training program with a new emphasis on the training of graduate students with strong mathematical and physical science backgrounds. 4.The continuation of the research and development program in computational molecular biology closely coupled experimental research laboratories, with a new emphasis on the integration of protein structural information into sequence functional analyses. A number of direction changes from the previous project periods are planned due to the changing nature and sophistication of molecular biology computational needs: l. a. Training in computational methods: To develop and provide training in the use and limitations of computational methods applicable to molecular biology, with a new emphasis on accessing, evaluating and using the wealth of network accessible computational and database services. b. Training in molecular biological applications: To expand the training of physical and computational scientists with strong analytical background in the analyses of molecular biological problems. 2. a. The development and evaluation of methods for function identification that integrate the available information obtained from various diagnostic pattern and database searches with other information such as protein structure, homolog family membership, genetic regulation signals, and enzymatic function(s) associations. b. The development and automation of new and existing methods for the exploitation of determined protein structure information in experimental design and analyses by non structural experts.